Compression of a Model to FP16

Model Optimizer can convert all floating-point weights to FP16 data type. The resulting IR is called compressed FP16 model.

To compress the model, use the --data_type option:

mo --input_model INPUT_MODEL --data_type FP16


Using --data_type FP32 will give no result and will not force FP32 precision in the model. If the model was FP16 it will have FP16 precision in IR as well.

The resulting model will occupy about twice as less space in the file system, but it may have some accuracy drop, although for the majority of models accuracy degradation is negligible. For details on how plugins handle compressed FP16 models refer to Working with devices page.


FP16 compression is sometimes used as initial step for INT8 quantization, please refer to Post-training optimization for more information about that.